Customer Reviews

Handbook of Statistical Analyses Using SAS, Second Edition

5 star:		(3)
4 star:		(3)
3 star:		(1)
2 star:		(1)
1 star:		(0)

 

 

This book is a good overview of how to implement in SAS some of the statistical methods you've probably already read about. I had done a lot of reading on factor analysis and clustering for my dissertation and needed a way to implement these methods in SAS. This book explains each method, then shows how to implement it using SAS. The data sets are from "The Handbook of Small Data Sets", which I got from my local university library. As I am a computer scientist and not a statistician, I found this book very helpful in putting together **the idea** of a particular statistical technique with the actual **implementation** of that technique. I have the first edition of this book also, and the second edition is much improved (especially the chapters on clustering and factor analysis).

Brian Everitt is the author of several very well-written statistical texts. Among them he has written a number that show how to implement statistical analyses usimg statistical software packages. This second edition of "A Handbook of Statistical Analyses using SAS" he has coauthored with Geoff Der.
As a SAS user, I find this book very handy along with other similar texts that I have on the use of SAS. What is particularly good about this book is that it serves as a guide to the use of various SAS procedures and also as an illustration of appropriate statistical approaches to real applications using SAS.

It starts out with a nice introduction to the SAS prrogramming language and its syntax and progresses through simple descriptive statistics to categorical data analysis to regression and analysis of variance and then on to more advanced topics, including survival analysis, logistic regression, generalized linear models,longitudinal data analysis, principle components, factor analysis and cluster analysis. Appendices provide SAS MACROs and SAS solutions to exercises in the text.

What is particularly good about this book, that may set it apart from some of the others, is the expert statistical advice about the implementation and interpretation of results in SAS. They provide excellent scholarly references to the statistical literature to support their advice. As an example, I particularly liked their discussion of Type I and Type III sum of squares in the analysis of variance. They give a clear explanation of what each means and when they are equivalent and when they are different. In addition, they present their own view as to which is the appropriate one to use in given situations and support their view with quotes from other researchers. Opposing positions are also mentioned and referenced.

This book is very well written. Each topic is presented with an interesting example, including discussion. None of the topics are presented in great depth, so for example, this is not the book to use for learning factor analysis, or cluster analysis, etc. The real strength of the book is that it shows how to do the analysis using SAS in a clear and concise way. I would recommend this book highly for anyone who would like to get started using SAS.

The authors covered many topics in applied statistics, but they didn't mention anything about time series analysis. I am disappointed after reading this book. The biggest problem with this book is that it's overly simplistic - typically only one technique is illustrated for each topic - for example, in cluster analysis, only hierarchical clustering was mentioned and there was nothing about partitional algorithm. The authors only used very small datasets, which ignored the biggest power of SAS, the ability to handle large datasets. The authors also printed all raw datasets in the book, which took quite a bit of space.

The authors should read Venables and Ripley's Modern Applied Statistics with SPlus first. Venables/Ripley made a great example on how to write an applied statistics book using a specific software.

I'm currently in a class that uses this book as a teaching text. It is a very handy instructive text- it's readable and informative. It's not for the beginner in statistics - you should have had some other formal training in statistical methods before this book becomes truly useful. I'm not new to statistics but I am definitely new to SAS and this has served as a fairly clear, informative approach to navigating the rather tetchy syntax of SAS coding. This is very much an introductory work however and only provides a very basic overview of the topic for the beginner.

The only thing keeping this book from getting a full five stars is the index. The index is completely worthless with major gaps in topic listings - if I want to look up the syntax to print a box plot there is no way to do so using the index. If a command is listed, then it is often listed under the example heading - i.e. certain procedures were demonstrated using the example data set "Sandflies" and those procedures would be in the index under "Sandflies" instead of under their own name. There are also no chapter headings at the top of the page but that is a minor nuisance compared to the index. This makes it very much an instructive text and not a reference text and weakens the otherwise excellent presentation of the book.

The example data sets can also be found at [...]

This is a good book if you know statistical analysis. Do you know what to use, when and where? If you do, this book is good because if you are going to use SAS for analysis you have got to know your statistics. What good would FORTRAN be if you didn't know Algebra? You need both.

This is a nice book if you know the subject from another book ! - otherwise the explanation is limited.

Uselfull for experienced people in the field. You are expected to know the subject early on. the book mostly provied an example for each of the subjects and explains them tersly. This wasen't what i expected